A Novel Approach for Improving Similarity Search Using SVM Classification Algorithm

The World Wide Web has transcended from a read-only to a read-write web. The problem of identifying important online or real life events from large textual document streams that are freely available on the World Wide Web is increasingly gaining popularity, given the flourishing of the social web. Earlier work used efficient algorithm for detecting all important events from a document stream through named entity recognition and topic modelling. In existing scenario, the learning algorithms are considered to identify and detect the unknown important topics from the disaster document. The problem found in this work is that the effective recognition of interesting named entities in previously unknown free text documents remains an open problem. The problem of finding similar literals with different meaning, or different literals that describe the same entity should be addressed. Hence to overcome all these issues the Support Vector Machine (SVM) algorithm is utilized in the current work. The SVM is used to create a model which helps to identify most random and important events from the given document. It is also used to recognize the high similarity sentences and semantic events from the specified document. The required information is retrieved based on the ranking events. The top most ranked events will identify the important event as well as produce the rich semantic meaning. Ranking SVM can be successfully applied for the task of finding and learning a similarity function for the event identification problem.